How to judge a review site: what testing methods, affiliate disclosures and update dates really tell you

Testing claims, affiliate disclosures and update dates each tell you something different about a review site. Here is how to read them, and an eight-step checklist you can apply anywhere, including here.

The Review NYU is not affiliated with New York University.

We do not currently earn a commission from any link on this page. If links ever become affiliate links, we will label them clearly.

Short answer. Judge a review site by the evidence it shows, not the confidence of its verdicts. Real testing leaves traces: a method you can read, measurements with units, original photos of the product in use, and a stated test period, all things Google’s reviewer guidance asks for. An affiliate disclosure should be clear and easy to find, which is what the FTC’s Endorsement Guides expect; it tells you the site earns from your purchase, not whether it tested anything. Update dates matter only if the content changed: new models, rechecked prices, a note on what’s new. Add named authors, clear ownership and visible drawbacks, and you have a five-minute checklist you can use on any site, including this one.

How we judged

This is a research-based guide. We did not run a lab or audit any review site for this article. We read Google Search Central’s guidance, “Write high quality reviews” (last updated 10 December 2025, per the page), and we used U.S. Federal Trade Commission pages on endorsements and on its consumer review rule. Several FTC pages blocked our automated reader, so for those we relied on the official search excerpts; we flag that in the limits section below.

The checklist here is a set of reading habits, not a scoring system. A site can pass every check and still be wrong about a product, and a small site can be honest without a lab.

You can and should apply the same checklist to The Review NYU. Many of our consumer posts are research-based roundups rather than hands-on tests, and we say so in a “How we judged” section like this one at the top of each post. If a post of ours doesn’t make that clear, treat it with the same caution you would anywhere else.

Signs a site actually tested the product

“We tested” is the most valuable phrase in a review and the easiest one to fake. What separates real testing from a rewritten spec sheet is evidence you can see.

A method you can read

Look for a methodology note on the article or a linked page that explains what was measured, with what equipment, and under what conditions. A good method explains how products were chosen for testing, too: if only products with affiliate programs made the shortlist, that shapes the result before any test runs.

Measurements, not adjectives

Google’s own reviewer guidance asks publishers to “share quantitative measurements about how something measures up in various categories of performance”. Numbers with units (minutes of battery life, decibels, grams, degrees) are harder to invent consistently across a whole site than words like “impressive” or “blazing.” Check that the numbers are consistent between the summary and the detailed results.

Photos and video of this unit, in use

The same Google page asks reviewers to provide evidence such as visuals, audio, or other links of their own experience. Original photos show a product on a real desk, in a real kitchen, often scuffed or mid-use. Manufacturer renders on white backgrounds tell you nothing about whether anyone handled the item.

Time with the product

Some faults only appear after weeks: a battery that fades, a hinge that loosens, a coating that peels. A review that says how long the product was used, and ideally returns later with a follow-up, carries more weight than first impressions written in an afternoon.

Comparison and drawbacks

Google’s guidance also asks reviewers to explain what sets a product apart from competitors, cover comparable options and discuss benefits and drawbacks based on their own original research. A tester who has used five robot vacuums can tell you why one is better; a writer who has used none can only repeat the box.

What an affiliate disclosure tells you, and what it doesn’t

Most review sites earn money when you click a link and buy. Google describes the model plainly: with affiliate links, if someone follows a link from a review and buys, the review’s creator is rewarded by the seller. That is not a scandal in itself. Hiding it is the problem.

What the FTC expects

In the United States, the FTC’s Endorsement Guides apply to reviewers who get paid by the companies whose products they discuss. The FTC’s plain-language guide, “FTC’s Endorsement Guides: What People Are Asking”, says disclosures need to be clear and conspicuous, and notes that when several affiliate links sit inside one product review, a single clear and conspicuous disclosure may be adequate. In practice, that means a reader should not have to hunt for it.

Where it should be

As a reader, expect the disclosure near the top of the article, before or close to the first buy link, in plain words such as “we earn a commission if you buy through these links.” A disclosure that appears only in a site-wide footer, a separate policy page or tiny grey text after the last product is a sign the site would rather you didn’t notice.

What it does tell you

  • The site has a financial reason to send you to a store.
  • It is following, at least on paper, the basic disclosure norm.
  • Its wording often reveals more: “we may earn a commission” is different from “this post is sponsored by the brand,” which means the maker paid for the article itself.

What it doesn’t tell you

  • Whether commission rates influenced which products were picked.
  • Whether the brand supplied free units, or let the site keep them.
  • Whether the products were tested at all.

A useful extra signal: Google suggests reviewers consider including links to multiple sellers so readers can buy where they prefer. A review that links to several retailers, including ones that may pay little or nothing, is acting more like a guide and less like a storefront.

Reading update dates properly

An “Updated” date at the top of a review can mean a full retest or a single changed comma. The date alone tells you very little; what changed tells you a lot.

Published vs updated

A published date says when the article first went up. An updated date says someone touched it later. Trustworthy sites show both, so you can see that a “2026” buying guide was first written in, say, 2023 and has been refreshed since.

What a real update changes

  • New models added and discontinued ones removed or labelled.
  • Prices and availability rechecked, with the check date stated.
  • New test results or long-term notes on older picks.
  • A short changelog or “what’s new” line explaining the change.

Google’s guidance asks reviewers to describe how a product has evolved from previous models or releases. A guide that still recommends a model replaced a year ago, under a fresh “updated this month” stamp, is showing you a date that was changed while the substance wasn’t.

Quick test

Click the top pick’s buy link. If the retailer shows a newer model, a very different price or “no longer available,” the update date is doing marketing work rather than editorial work. For more on judging prices, see our guide on how to tell if a sale is a real deal.

Who wrote it, and who owns the site

Google asks reviewers to demonstrate that they are knowledgeable about what they are reviewing. As a reader, you can check that knowledge directly.

  • Named author. A real name with a bio that explains relevant experience. “Staff” or “Admin” on every post is a weak signal.
  • Track record. An author page listing other work in the same field. A person who reviews mattresses, VPNs, baby car seats and crypto exchanges in the same week is probably not testing any of them.
  • Ownership. An About page that names the company or person behind the site. Some review sites are owned by media groups, some by marketing firms and some by retailers; ownership tells you whose interests sit in the background.
  • Contact and corrections. A working contact route and a visible way corrections are handled. Sites that never admit an error are not sites that never make them.

Conflicts of interest beyond affiliate links

Affiliate commissions are the most common conflict, but not the only one. A careful site discloses all of these:

  • Free or loaned products. Whether the brand supplied the item, and whether it was returned.
  • Sponsored content. Articles paid for by a brand should be labelled as such, not presented as independent reviews.
  • Ownership ties. A review site owned by, or partnered with, a company whose products it ranks.
  • Ad deals. Display advertising from a brand on the same page that reviews that brand.

The FTC has also moved against manipulated reviews more broadly. Its final rule on fake reviews and testimonials, announced on 14 August 2024, covers fake or false consumer and celebrity testimonials, buying positive or negative reviews, and insider reviews. The FTC’s questions-and-answers page on the rule says it lets courts impose civil penalties for knowing violations, and a December 2025 FTC business blog post about warning letters puts those penalties at up to $53,088 per violation. That rule targets reviews and testimonials rather than every editorial roundup, but it shows the kind of behaviour regulators consider deceptive. For star ratings on store pages, see how to spot fake online reviews.

Red flags and what they usually mean

No single item below proves a site is untrustworthy. Several together are a strong signal.

Red flagWhat it usually meansWhat to do
“We tested” with no method, photos or numbersThe claim may be marketing languageLook for an independent review with measurements
Only manufacturer imagesNobody may have handled the productTreat it as a research summary, not a test
Disclosure only in footer or policy pageCommercial ties are being played downAssume every buy link pays the site
Fresh “updated” date, outdated modelsDate changed, content didn’tCheck the top pick at a retailer
Every product rated 4.5 stars or higherLittle real comparisonLook for a source that names losers
No drawbacks for the top pickSales copy, not reviewSearch for owner complaints and known faults
Author is “Admin” or has no bioNo accountable expertiseGive the article less weight
One retailer link only, on every pickPicks may follow the best-paying programCompare prices at other stores
Countdown timers or “deal ends tonight” bannersPressure, not guidanceSlow down and verify the price history

An eight-step checklist for any review site

This takes about five minutes per article once you know where to look.

  1. Find the method. Look for a “how we test” or “how we judged” note on the article itself, or a linked methodology page. If neither exists, assume the post is research-based or opinion until shown otherwise.
  2. Check for the reviewer’s own evidence. Look for original photos or video of the product in use, measurements with units, and a stated test period. Stock press images and spec lists copied from the maker are not evidence of testing.
  3. Find the money. Locate the affiliate or sponsorship disclosure. Note whether it sits near the top and near the links, or only in a footer or separate policy page.
  4. Read the dates. Note the published date and the updated date, then look for a line that says what changed. Check whether prices, models and availability match what retailers show today.
  5. Identify the people. Check for a named author with a bio, an editor, an About page that names the owner, and a working contact or corrections route.
  6. Look for the downsides. A trustworthy review names drawbacks and says who should skip the product. Count how many picks get any real criticism.
  7. Compare the links. See whether the article links to more than one seller and whether its top pick happens to be the one with the only buy button.
  8. Cross-check one claim. Pick one specific claim, such as battery life or a measured result, and see whether another independent reviewer or the maker’s own spec page agrees.

If you write reviews yourself, the same points work in reverse; our guide to writing a helpful product review covers them from the writer’s side.

Research-based roundups: useful, if labelled

Not every useful review is a hands-on test. A well-sourced roundup that reads professional test results, manufacturer documentation, return policies and owner reports can help you narrow a shortlist, especially for expensive or rarely tested products.

The test is honesty about method. A research-based article should say so near the top, name its sources, link to them, and avoid language that implies first-hand use. It should not include star ratings or “our testing found” phrasing. When a research roundup and a hands-on test disagree on a measurable point, give the measurement more weight.

That is the standard The Review NYU tries to meet, and readers are entitled to check whether we do. If one of our posts implies testing it doesn’t describe, that is a fair complaint to raise.

Bottom line

Trust a review site in proportion to the evidence it shows: a readable method, its own photos and numbers, a disclosure you can’t miss, update notes that say what changed, and named people who can be held to account. Affiliate links are normal; hiding them is not. Dates are useful only when the content behind them changed too. Run the eight checks on any site, ours included, and give the most weight to sources that measure things and admit drawbacks.

Limits of this article

  • We did not test or audit any review site for this article; it is based on reading official guidance.
  • The FTC pages we cite returned an access error to our automated reader. FTC statements here come from official search excerpts only: the clear-and-conspicuous standard and the single-disclosure note (Endorsement Guides Q&A), the categories covered by the fake reviews rule (14 August 2024 press release), civil penalties for knowing violations (rule Q&A) and the up-to-$53,088-per-violation figure (December 2025 blog). Read the full FTC pages before relying on them.
  • Advice on where a disclosure should sit (near the top, near the links) is our reading of the clear-and-conspicuous standard, not a quotation of a specific FTC rule.
  • FTC rules apply in the United States; other countries have their own advertising disclosure rules, which we did not review.
  • Google’s reviewer guidance describes what Google wants from publishers; it is not a guarantee that pages following it are accurate.
  • We did not read any specific publisher’s testing methodology page for this article, so we describe good methodology in general terms rather than naming examples.
  • No product images are shown: we did not have properly licensed photos, so we left images out rather than use placeholders.

Sources

Frequently asked questions

Does an affiliate link mean a review is biased?

Not automatically. It means the site earns money if you buy, which creates an incentive; look at whether the review names drawbacks, compares rivals and links to more than one seller.

Where should an affiliate disclosure appear?

The FTC asks for disclosures that are clear and conspicuous. As a reader, expect it near the top of the article and close to the buy links, not only in a footer or a separate policy page.

How can I tell if a site really tested a product?

Look for a method, original photos or video of the product in use, measurements with units and a stated testing period. Stock images and spec lists copied from the maker are not evidence of testing.

What does an updated date on a review mean?

Only that someone edited the page. A meaningful update adds or removes models, rechecks prices and says what changed; a new date on an old list of discontinued products is a warning sign.

Are research-based reviews worthless?

No. A clearly labelled roundup that cites professional tests and official pages can help you build a shortlist. It becomes a problem when it implies hands-on testing that never happened.

Does The Review NYU test products hands-on?

Many of our consumer posts are research-based roundups rather than hands-on tests, and the How we judged section at the top of each post says which kind it is. Apply the same checklist to us.